PEPFAR Quality Control and Method Validation Activity 14: Part 5

2 (00 Slcpe(m) companson Companun methods comparative Compansct) companison companson Comparison Compatismn 90 OF m€#cds 65 paten mpatison 2 (oc) 3 60 35 Reference Interval Experiment Reference (00 comparison 50 2 loc) 62 Process for Introducing a New Method at Gaussian Distribution is the Key loc) Questionnaire (00 patison pacism Abs Compar€n [00 2 loc) Documentation of Companzson Documentation of Laboratory Summary Sheet Summary Sumnmary Sheet Companson Reference Interval Referenco Why intorvnåaluationp Sl$e(m) Crcating A bsolu+e loc) compantson lox) Companson Companson OF comparison OF methods OF compansmn Method Validation Plan PCI Qthod Validation+i Compansm 205 Why Perform Beterencc mpariso comparison of methods zoo Alosolu±e 205 S ISO 15189:2012 - 5.5 Examination processes 15189:2012 - 5.5 Examination Comparison OF methods 15189:2012 - 5.5 processes loc) 'Judge the acceptability of observed b'öecence comparison Compactson 205 parison zo 60 -zoo 2 (00 -zoo 60 zoo 205 S zoo zoo Slcpe(m) Absolu+e 905 lotioductioninto RoutiÖe Service Sl«e(m) Sløe(m) DFFerence zoo loo Questionnaire Documentation ot laboratorv Documentation ot laboratory Introduction into Routine Service Creating the Questionnaire •Collect expenrimental data •CoLlect experimental data •Cohect expel Irnental data •CoDlect experimental data •CD'@ct experi'mental data •CoHlect experimental data •Coblect experimental data •Conlect exper•mental data •Contect expenimental data •Con:ect experimental data • Collect data •Conlect expertmental data •Conlect expenimentÄI data • CoLl'ect experimental data experimental data 10 •Collect experimental data • exper•mental data •Conlect experimental data 205 00 ISO 15189: 2012 - 5.5S Examination processes ISO 15189:2012 - S.S Examination processes ISO 1518922012 - S.S Examination processes 60 ISO 2012 -S S •Collect data 55 50 35 S19e(m) your Site Abs Sltg(m) / vuol. 5.5.1 Selection, verification and validation of examination Summary Sheet 5.5,1 Selection, verification and validatio loc) Director's ADDrovaI of the New Method Director's Approval of the N Director's Approval of the New Method performance characteristics S.S. 1 Selection, verification and validation of exam'ination 5.5. 1 Selection, verification and validation of examination S.S. 1 Selection, verification and validation of to Statistical Quality Control Puality Director's Approval ot the New Method Director's Aooroval ot the New Method Director's Amjroval ot the New Method New Abs effods Should Include contact •n case the Should include contact the Should Include contact Inforrnatron In the Should include contact Inforrnatlon In case the Should Include contact Inforrnat'on In case the Should Include contact 'n case the • Should include In the Should include contact Inforrnatlon •n case the Should include contact •n case the Should include contact In the Should Include contact In the Should Include contact Inforrnatlon In case the Should Include contact •n the Should Include contact In case the Should include contact Inforrnatlon the Should include contact In case the Should Include contact Inforrnatlon •n case the Should Include contact Inforrnatlon •n the Should include contact informnation In case the Should include contact 'n case the Even though results are accurate and precise, • Should In Should contact Bbs Should include contact information in case the Should Include contact Infornatlon In case the Should include contact informnation in case the /buo/ •Use stat'st•cal tools on the data to •Use tools on the data to • Use tools on the data to •Use statistical tools on the data to •Use stat.st.cal tools on the data to •Use stat.st•cal tools on the data to estimate •Use tools on the data to estitvate • Use statistical tools on the data to est)mate •Use stat.st'Ql tools on the data to •Use tools on the data to estuvate •Use statistical tools on the data to esttntate •Use stat.st•cal tools on the data to estrnnate • Use stat.st•caf tools on the data to estitrvate • Use stat.st•cal tools on the data to • Use stat.st:c-al tools on the data to •Use stat.st:cal tools on the data to est.:vate •Use tools on the data to estimate • Use stat.st•cat tools on the data to •Use stat.st•cal tools on the data to esti:vate •Use stat.st•cat tools on the data to • Use tools on the data to estirvate •Use stat.st•cal tools on the data to •Use st3t.st•cal tools on the data to •Use stat•st'cal tools on the data to est.nnate •Use stat'st•cal tools on the data to estimate •Use stat.stncal tools on the data to est"vnate • Use stat.st•cal tools on the dota to Y. X. zoo Partition P artlt'an Par titian Parttt'on procedures E•clusion Criteria Enclusion Criteria Partition Criteria Exclusion Criteria Criteria E•cluslon 205 Abs 205 Abs Abs 205 • Establish a reference interval Establish a reference interval for a loo expeninnent reveals sonw petential e•perwvent rewals sonre potential expert nnent reveals potential experrnnent reveals sonwe potential e•petlnnent re• eals potential expennnent sonw potential expertn7ent reveals potential evperirnent potential expennnent reveals potential expenn7ent sonw potential experinnent reveals sonw potential abnor experin7ent reveals potential €»pennnent potential expennnent potential experinvent reveals sonw.e potential eyperlnnent reveals potential expennnent potent 1.11 experinnent potential experirvent reveals sonw potential experinnent reveals sonw.e potential expert 'Vent reveals potential expenruvment reveals sonw potential experin7ent soneale potential experinnent reveals sonw potenwial experimvent reveals potential Abs reveals potential reveals sonw potential sonw potential potential zoo experi'vent reveals potential experinnent re•veals sonw potential expeninnent reveals some potential exper:nnent potentital experitvent potential experin7ent sonw potential experinnent reveals sonw potenttal expennnent reveals sonw t»penment .1hn reveals abnor reveals sonw experin7ent reveals some experitvent reveals experinnent reveals sonw experttvent reveals sonw e•perlnnent reveals expennnent reveals experinnent reveals reveals some noun mvn mwn rovn Opertrvent expennnent e•perlnnent euperltvent experinnent some expeninnent of errors of analytical errors of analvt'c 31 errors of of analvt'cal1 errors of analytic .31 errors of analvt•cal of analvt•cai errors of analvtlcal errors of analßtacal errors of anal' tic al of analvtjcal errors of anal', tic al of anal' ICal of analvt•cal errors of analytical of analy tic 31 of analy tical1 errors oc sonw zoo loo) zoo ethods -zoo 205 bFerence AST 2 IOC) 2 loo Difference b"FFecence Exclusion Criteria E•clusion Criteria procedures Partition Criteria E•clusion Criteria Parti tion C 60 zoo bFerence Describes the variations of a ' Describes the variations of a ence intended Intended Define quality requirement • Define quality requirement concentration %Bias or Sigma AST at Long-term Total 90 b'öecence zoo blF$ecence ence 5.5. 1.1 General S.S.I.I General S. S. n.1 General S.S. i A General General •Compare the With the defined •Com;nre the With the defined •Como are the the defined the the defined the defined •Comcare the the •Comoare the •Ccmoore tine With •Comoye With the denned •Compare the the defined •Compare the With the •Compare t6e With the denned •Compare the •Comoare the the def ined •Compare the defined •Compare •Como are Withe the defined •Como are the •Com;Jare the With the •Compare With the •Compare the errors With the •Comcare the With the defined •Compore the With the • Comoare the the •Comoare the the denned • Compare the denned • Compare With the •Compote the With the •Compare the With the denned •Compare the Withe •Compare With defined •Compare the errors the defined •Compare the the •Como are the the •Compare the With the denined •Comoare the errors the •Compare the With denned •Comcare the the defined • Como are the the defined • Comcare the With •Comoare the With the de• •Com;nre the the •Comoare With •Comoare the w:th • Compare the the defined the With the errors With the the the the With the the With defined tne With the denned With reported results will be clinically misleading if the reported results will be clinic reported results will be clinically isleadlngÆfth the tne if the experiment reveals some potential abnormalities nti use Ys Sty) Abs CharodETisf'«s Ckarod?fisfks Ckarodensfks Ckarodensi'rs Charodenst.cs Pre-purchase Assessment 20 Abs OC ompafison o &hods of Xc %difference 120 Concentration . Precision Partcition Precision Error Metric Abs % bifecence Cocrelafion measurement procedure and reference measurement procedure and Chauct ris€lcs o ref C at error 205 ompa€tson o ethods o ethods Abs ods valåe in healt y Characteristics of the :rence s, loo 90 PEPFAR QC Workshop Abs ompa€tson o &hods ence ompafison o efhods o effods ethods met-hods et-hods ompafison o ompaqÅson o ompaqlson o böecence tot Mav be consent be consent to' Site's • Mav consent Mav be . Mav be be . be be , consent reference interval does not match our o ulatiån Details about the candidate at 9fXc Site's OC •vas oc Of of Xc Select QC procedures • Select QC procedures with Abs 35 35 Regression s, 2 loo Bias oc Site's with Site 2 (00 bFecence 90TE measurement or value in healthy Bias 50 'e oc s, BAS reference interval does not match your population reference interval does not match your po ulation SD With (00 Abs Of measurement or 120 loo) el afion Reference The laboratory should have a mechanism in place for TEA Accep DFecence 35 _ A ccefb Accep 35 comparative selected reference 35 zoo effods 2 loo comp BFecence population. ence 5.5.1.1 General CocreIaGon Slope(m) C) S lop •on reference individual that, if Select the appropriate types of experiments to reveal analytical errors • Select the appropriate types of experiments to reveal analytical errors TEA m ints t SlQpe (m) C) Xc Slope (m) C) Bias TEA methods ocrelafion C) AST y) 35 % Verified Abs biFFecence Adult 'If method is acceptable, then Accep 2<30 2<90 method ao lox Can Methods Y-.n+eccep+ ethods Bias oc ence 3 2<90 35 Method Approval AST • Judge the of •Judge the of o •Judge the -f blFFecence •Judge the observed böecence Aca •Judge the • Judge the •judge the •Judge the of •Write g pr ceder / • Write operating procedures Write operating procedures AST 2 loo (yes OC Installation and Calibration 90) Sl.e(m) TEA 35 Y-tn+ercep+ Y-•n+eccep+ Y-jn+eccep+ b' n+eccep+ Ya.n+eccep+ n+eccep+ 90 Slope (m) individual that divide the Q) medical review and confidential notification 2 loo AST 62 Reference or OC Pe ao present, serve to keep that •ve to keep that individuals. Slope performance Can The laboratory shall select examination procedures which la ratory se\ect- xamination p 'Y ethods Can CAn TEA • 120 samples for each partition 120 samples for each partition TEA Accep ov of that ( yes 35 tore Ional Cons+on+ Conston+ Cons±on+ cons±on+ (yes Accep TEA 1 loo -in ional (Expected regression tesress.on Analyte Accel Can - In conston+ 3 2<30 (yes or Can (yes (yes (Yes reference sample into referenc (Oto yes or on o &hods method s, Lirear 2<30 ao 90 yes individual from being i„orn being 2 loo lox zoo perform reference interval Collect experimental data ' Collect experimental data •Collect experimental data • Collect experimental data oc OC 102 • Train analytical staff Train an ta f ana?yticai staff 120 iQS or Can Range Cited Can have been validated for their intended use. 35 May be accompanied by a consent form Whyn to Perform Reference Interval Evaluation? When to Perform Reference Interval Evaluation? Whcn Perform Referencc Approved / Not Approved Lirear. S (syr) (yes s (syr) S(sy) Lir-ear Linear. or (yes yes or se TEA Verify a reference interval for a mae Verify a reference interval for a measurement Sty) oc Can 120 Activity 14: 120 120 (00 Can significant subclasses _asu % C IQS or TEA 290%/0) yes Accep included in the reference es as a basis for a provider to Serves as a basis for a provider to Familiarization Period 2 loo (00 Accep or Ian AST Rearesston ?eéormnce '110 •IQS OC 120 % c vs S BAs S of % c S BAS Notify your customers if sarnpl}röhection y) % C Regression y) ) S 's Can Regression Regression y Lirear. y) • Notify your customers if sample collection • Notify your c,ustomers if sample collection Xc at Xc Reoression y Regress\on AST procedure and reference popula iöh. lox ae Xc 2 loo Rea resson or NT PE NT Re3ressxon at Xc Xc PE C N NT reeres.on s: Example: "Cape Clinic Laboratory personnel are at Xc Xc % c NT af Xc Can 120 experiment V AST procedure and reference population. V AST AST 35 U/L 3.1% AST 6.3% 3.1% 15.6% 4.4 Acceptability (-s€/xc - vcs y'-Xe Examples data collection. Lirear. lo Use statistical tools on the data to estimate size of analytical errors • Use statistical tools on the data to estimat_e size of analytical errors TEA ct Xc y'-Xc 120 v'-Xc S of 35 If not approved, provide recommendations/corrective actions below. y'-Xc % gtas or % or 120 %8as or %g•tas or %ßas or %åas or %ßas or % g•tas or Zcv XYZ Bias oc or T om or —r or % C V AST Zcv % •Jas C Zcv ZCV or or ao 2.2 Cons PE C NT NT V AST C NT eE C NT AST • New analyte 90TE 90 90 T E or % c 90 TE yo TE or (yes S BAS IQS % c : 120 interpret a set of results for a of results for a AST % Bias or Xc 120 IQS y'-xc cob ao OC 35 oc %8as or Bias or %gtas or Ac %ßas or or Xc allowed to obtain specimens and use the laboratory C NT % or S Bas % c AST %g•tas or % Bras or S BAS S BAs PE % c requirements or reference interval have 120 Introduction to Method Evaluation % C e c NT PE C NT intended — Gender 120 UJ/L 120 U/L O U/L or 6.3% 62 10.8 Acceptability Examples NT y'-Xc (.s€/rc 20 samples for each partition rethoe or T AST V AST af Xc loc) 35 s, %8as or 90 TE PE ave Xc , 35 y'-Yc V AST Method Validation Plan or PE C NT Accep (J 13 % c 90TE 63 oc Zcv %gtas or i 35 NT %gtas or exe Manufacturer Accep %ßas or oc Zcv AST ZCV % g•tas C0bas 9b C 90gtas or 35 20 y'-Xc or 35 35 35 Z 20 AST 205 120 21) The specified requirements (performance specifications) for The specifie d requirements (performance specifications) for AST oc or 65 STE or ZCV % gus or PE C NT 90gtas or 90TE ) S IQS loc) Roche oc or Ac Laboratory Director: Human af Xc 90 TE Xc % c V AST V AST Compare the observed errors with the defined allowable error • Compare the observed errors with the defined allowable error or TEA % gtas or 90 TE values and questionnaire information for the — Age Xc %ßas or Accep Cobas c501 y'-Xc — Alcohol consumption Change instruments or methods @'öttgntents or methods Ac particular patient. Roche or Se€um Sesum Secom 35 Test Method AST 120 ZCV Roche Zcv ZCV • Transfer a reference interval from a comparable Transfer a reference interval from a comparable ' Transfer a reference interval from a comparable changed 10 Roc ho 35 AST Method %8as or Roche Cobas Cobas c501 V AST 35 Ac Insert Lab director name here or 1000/0 2900/0) Females Ranges Roche Co bas Roche Cobas each examination procedure shall relate to the intended each examination procedure shaii relate to the intended each examinration procedure shall relate to the intended 55 35 35 squ — Race Race Validation Studies 65 Test Method • Intr duce method into service ("do-live Test Method Test M Test M ethod determination of reference intervals. All information — Recent illness AST Method 35 PE C NT XYZ Analyzer C NT Part 5 Test Method Validation Studies (Y) 35 Human 'Introduction into Zcv 110 Validation Studies Cobas measurement procedure. measurement procedure and r AST XYZ Analyzer XYZ Analtyzer xyz S Goa • Introduce method into service ("Go-live 35 use of that examination. — Stage of menstrual cycle (Y) measuremen XYZ Ana Change sample collection systems xyz XYZ %g•tas or or Method Comparative % Bas or Sec um AST Metho oc Roche or Ocho Sometimes called a "reference range" ' Sometimes called a "reference range" AST • Judge the acceptability of observed performance characteristics Judge the acceptability of observed performance characteristics — Pregnancy AST % c will be held in a confidential manner. Notification of Zcv Comparative Method (X) Method PE C NT Prepared by: Date: o tidation Stu Method or Cobas Met hod 110 35 Cobas c504 90 TE 0000 AST — Stage of pregnancy Moth o 120 % ias c Use linear regression equation • Use linear regression equation Use linear reg *ion equation 5as or %åas or — Recent Surgery Roche Cobas Insert name and title here date") the result will only occur if the laboratory director esult will result Roche Cobas Test Method AST Method Roche Cobas Validation Studies • Reformulation of reagent rest Method Test Method or "normal range" (old term) Validation Test Method — Blood group 35 Routine Service (Y) Test Method — Recent blonod transfusion — Recent blood transfusion If method is acceptable, then perform reference range experiment If mnethod is acceptable, then perform reference range experiment If method is acceptable tn C NT (Y) PE C NT • Limited to one time transference Limited to one time transference (Y) or (Y) CObas deems the result requires further follow-up." deems the result (Y) cobas deems the resul deems the r res further ollow-up. res further res further o ow-up. Roche Cobas Roche Cobas Roche % or or cr (Y) • Monitor routine performance Yoni rot,ltine performance — Ethnic background ZCV % Bas or C0bas C NT Roche cobas — Abnormal blood pressure Test Method Test Method PE C NT tae ROC he 66 62 65 6% 70 63 68 60 Roche 72 67 76 75 35 CObas c501 Roche Roc he (Y) From Westgard.com Website. The Experimental Plan for Method Validation. Last accessed May 1, 2013 From Westgard.com Website. The Experimental Plon for Method Validation. Last accessed May I, 2013 Cobas c501 Roche Cobas or Cobas c501 C0bas c501 Test Method Cobas CSOI CObas c501 cobas c-501 Roche c501 Cobas c501 Cobas c50f Cobas Cobas c5nq AST